Fire risk posed by solar panels

[Pascals Wager]

Originally Posted by hellosailor

So in theory, placing a fuse immediately adjacent to the panel would open the circuit and prevent a fire from a shorted panel. Presuming, that the "short" was drawing enough power to throw it. I'm not clear on whether that would still allow hot spots, in a working/rated service, to still run hot enough to start a fire. From what I'd heard, they can still start a fire while within normal current range. Not? For sure?

And that the main risk of fire hazard really is improperly built panels, i.e., the cheap ones without internal bypass diodes?

The fuse would only protect the wire, which is not the fire source. A dodgy panel may generate enough resistance to catch fire but the current can be below any fuse as you suggest.

[hellosailor]

Pascals-
No, the fuse might protect the PANEL as much as the wiring. Consider, if that panel has a 20A rated output and you have a 20A fuse next to it. As soon as the "short" on the panel starts to actually consume power--the fuse blows and protects the panel, presumably before it can reach combustion. As I'm hearing it, the problem is that the short circuit on the panel is what is creating the overload, and a fuse next to the panel will end that by breaking the circuit. If the short is entirely within the panel and not dependent on the rest of the circuitry...that's something else. Totally unclear which is the case here.

[nilespf]

"As soon as the "short" on the panel starts to actually consume power--the fuse blows and protects the panel, presumably before it can reach combustion."


If the panel is direct wired to a battery then the battery can source enough current to burn everything in the path from the battery to the point of the short.



If the panel is part of a system of multiple panels wired into the input of a MPPT and the panels are wired in parallel then even though the output voltage will drop if not go to zero, there still is the possibility the other good panels can source their individual Isc into the so called shorted panel.



If the panels are wired in series then Isc is no longer additive and there is no source for obscene levels of current.



I haven't read anything as to be able to comment if most fires are the result of a panel becoming a consumer of current i.e. a short OR if a PV cell can degrade such that it no longer produces rated current.



My guess is (assuming no trauma to the panel) if a PV cell ceases optimum current production and has a subtle increase in conduction resistance then we know the increased voltage drop may not be high enough to trigger the bypass diode into conduction but the added resistance does become a heater (at the current levels we operate at) elevating localized temperature. It doesn't take much in resistance to start the heater process and because the PV cell is still conducting all the other down stream cells are still producing further ensuring the bypass diode doesn't turn on. Now if you shadow a cell all bets off and the cell operating resistance increases dramatically. With 10A panel current trying to pinch through a high resistance, the voltage drop immediately slams the bypass diode on. I believe most bypass diodes are located in the J boxes and are soldered by hand.



Find the data sheet on the Q6LPT3-G3 PV cell. I have used this for modeling and it agrees very well with my Renogy panels.


Sure hope we can get our hands on statistics about these fires.


Phil

[Pelagic]

Interesting stuff that may affect how I rewire my Solar panels after finishing the fiberglass Bimini which seals out the rain between the panels


I'm lucky, the 4 Sunpower panels are open from underneath, so would get maximum cooling and monitoring.



But it is the Series/Parallel from a Safety aspect that I am having a layman's problem understanding ?
The rated voltage on my 4 Sunpower 327 w panels is
Vmpp 54.7V

If I connected each pair in series it would be 109.4V.
Is that Safer?


I have a 45a 4pole breaker between Panels and Outback 80 and a 125a 4pole breaker between Outback 80 and house bank
House Bank has a 400a Fuse

I was originally parallel connected, but when Bimini is finished, there will be a combiner under aft panel and a shorter exposed wiring run, so should I stay Parallel or go Series?

[nilespf]

Pelagic,
I was looking at the Outback 80 and it wants a Voc less than 150 V. If I read the specs correctly, your panels in series can stuff up to 260 Voc into the controller (when all are in series). I had the same problem in my setup i.e. 4 panels but my controllers I suspect are infinitely less expensive so I chose to use two with each handling a pair. My controllers also have about the same input voltage specs but I noted outback wears that as a badge of honor. I never saw it as a limitation but rather as a challenge in system configuration. It does indeed force a configuration less sensitive to shading. One may consider two in series in parallel with the other two in series into a MP4 Y connector. I noticed your panels have 3 diodes so you won't lose the entire panel. The goes outta of the Y branch MP4 will have to be sized a bit different. It will keep the currents down but provide a lift (relatively speaking) if one panel on each string goes dark.


I wonder what the real odds are that a panel can become a sink for current i.e. a short. I hear about these fires and w/o panel trauma I cannot imagine how a cell suddenly become a short circuit. Some talk about box fires which can arguably point to a bypass diode not properly soldered into place. I don't know but my panels do have an air gap above the Bimini but not much. I'm going to spend more time looking at them.



Phil

[a64pilot]

To my knowledge all the boat fires have been from flexible panels.
There have been several almost fires as well of course when the panels burned through, but the panel wasn’t on anything that would burn, so no fire.

My panels are fused too, but it’s not going to help prevent a fire, the cause of the fire appears to be an internal short in the panel, if so when the short occurs the panel output drops cause the power goes into the short, a fuse won’t help.
It’s not current flowing backwards from the bank into the panel, cause I believe that isn’t possible due to the diodes.

[CptCrunchie]

https://www.renogy.com/voluntary-product-recall/

[nilespf]

CptCrunchie, another nail in the coffin for flexibles

[Pelagic]

Originally Posted by CptCrunchie

https://www.renogy.com/voluntary-product-recall/

My guess is that this is only being communicated in the USA and other progressive countries to avoid Federal Safety Investigations.
These panels will end up being sold in 3rd world countries for simple families trying to charge thier cellphones

[nebster]

Originally Posted by sailorboy1

Since I wonered it and have learned over the years that when you go "someone such made XXX" that someone probably does. I Goggled it and yes they made MC4 connection fuse holders that go right in line. For the $20 I think I'll get some and have the just in case protection.

Yup, I have MC4 inline fuses on all of my strings, too.

For clarity: these fuses are to protect the wires in the event of a backfeed from the paralleled strings. They won't help with a thermal issue due to a fault inside a single panel membrane, which I suspect is what's happening most of the time with those flex panels.

[Maine Sail]

Originally Posted by Pelagic

My guess is that this is only being communicated in the USA and other progressive countries to avoid Federal Safety Investigations.
These panels will end up being sold in 3rd world countries for simple families trying to charge thier cellphones

Or on eBay or through other less reputable buying venues. If a company the size of Renogy can't get a handle on the quality out of their contracted Chinese supplier how is a boat owner to know what is good or safe when sourcing semi-flexible panels off eBay, AliX, Amazon etc... Scary stuff for sure..

In the discussion around fires we should also not forget counterfeit/knock-off MC4 connectors. Every MC4 failure we have seen has been a cheap Chinese knock-off. We have yet to see a single failure of genuine Stäubli/Multi-Contact branded MC4 terminal.